Abstract— In everyday transit, the issue of blind spot collisions are a prevalent issue. Currently, there are blind spot detection systems available as factory accessories to manufacturer specific vehicles, but aftermarket detection systems are scarce. Blind spot detection systems all share a common feedback to the driver; visual and/or audio. In this paper, a haptic feedback steering wheel cover will be used in conjunction with ultrasonic sensors to monitor blind spots. This will provide a better alternative to a blind spot alert systems by adding haptic feedback. A preliminary study was performed on human subjects to analyze their response times to different haptic alerts to ensure the most beneficial stimuli is used on the blind spot alert system. However, different modes of vibration showed no significant difference in response time, so the stimuli was chosen in terms of user preference. Subjects explained that they felt a slow pulsing vibration was more effective and distinguishable than a steady vibration. I. INTRODUCTION Blind spots are a common nuisance that drivers encounter in every day travel. According to the US Bureau of Transportation Statistics, there are 254.4 million registered passenger vehicles in the United States and approximately 840,000 blind spot collisions occur with 300 fatalities every year [1]. Blind spots are defined as the areas in adjacent lanes of traffic that are blocked by various structures in the automobile or other indirect areas not visible to the driver [2]. Such physical obstructions can potentially cause a collision if a driver performs a lane change when another vehicle is present within the blind spot. To prevent this from happening, some manufacturers have incorporated blind spot monitoring systems in their vehicles, i.e. Mercedes Benz (BSM), but these systems are factory accessories only installed into their vehicles and not available as an aftermarket product. These blind spot monitoring systems warn the driver by illuminating a light and/or producing an acoustic alert. The driver's visual attention to the road is vital in preventing collisions. Just for perspective, if a driver is traveling at 70 mph and takes their eyes off the road for 2 seconds, they would have traveled 205 ft unattended [2]. By using a warning light, it would require the driver to take their eyes of the road, so the driver is not aware of the environment in front of them. In order to help prevent blind spot collisions, a haptic blind spot alert system has been produced at the University of South Florida, Tampa, FL, USA. The system utilizes an aftermarket steering wheel cover with multiple vibrations motors, an arduino microcontroller, and two ultrasonic sensors. The alert system detects whether a vehicle is within the blind spot when the turn signal is engaged. If there is a vehicle present, a haptic vibration is sent to the driver through the steering wheel cover. Related products for blind spot detection systems and haptic avoidance systems are listed in Section II. The hardware setup and coding are covered in Section III. The experimental setup and procedure is covered in detail in Section IV. The study results are explained in Section V and finally, the conclusions and future recommendations will be covered in Section VI and VII. II. RELATED PRODUCTS AND SYSTEMS There are multiple products out today that provide blind spot detection. Volvo offers the "Blind Spot Information System (BLIS)" on their vehicles as an accessory [3]. BLIS utilizes a digital camera to monitor the blind spot region and illuminates a symbol on the mirror housing. Mercedes Benz offers "Blind Spot Assist" also as an accessory on their S and CL-Class vehicles [4]. This system uses a short range radar systems and alerts the driver also by illuminating a symbol located on the side mirror. Toyota also has a blind spot detection system (BSM - Blind Spot Monitor) that works on the same principle as Mercedes and Volvo; illumination of a symbol on the side mirrors. This requires the driver not only to check their side mirrors for approaching vehicles, but also to check for the blind spot symbol. Figure 1 is an image provided by Toyota of the blind spot monitor alert on a Toyota Camry. Figure 1 Toyota Blind Spot Alert Select automotive companies offer haptic alerts in some of their vehicles, but none of these systems are for blind spot detection. Mercedes Benz offers a system called "Lane Keep Assist", which monitors road markings for any deviation past a solid line. If the vehicle passes over a solid line, it provides a vibration to the steering wheel to alert the driver[4]. This vibration feature is incorporated into the steering wheel, but is not available for "Blind Spot Assist". Haptic Blind Spot Alert System Christopher Nebolsky, Mechanical Engineering Student, University of South Florida
16
Embed
Haptic Blind Spot Alert System - REED Lab at USFreedlab.eng.usf.edu/photos/HapticsDemos2013/2013Nebolsky.pdf · approaching vehicles, but also to check for the blind spot symbol.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Abstract— In everyday transit, the issue of blind spot
collisions are a prevalent issue. Currently, there are blind
spot detection systems available as factory accessories to
manufacturer specific vehicles, but aftermarket detection
systems are scarce. Blind spot detection systems all share a
common feedback to the driver; visual and/or audio. In this
paper, a haptic feedback steering wheel cover will be used in
conjunction with ultrasonic sensors to monitor blind spots.
This will provide a better alternative to a blind spot alert
systems by adding haptic feedback. A preliminary study was
performed on human subjects to analyze their response
times to different haptic alerts to ensure the most beneficial
stimuli is used on the blind spot alert system. However,
different modes of vibration showed no significant difference
in response time, so the stimuli was chosen in terms of user
preference. Subjects explained that they felt a slow pulsing
vibration was more effective and distinguishable than a
steady vibration.
I. INTRODUCTION
Blind spots are a common nuisance that drivers encounter
in every day travel. According to the US Bureau of
Transportation Statistics, there are 254.4 million registered
passenger vehicles in the United States and approximately
840,000 blind spot collisions occur with 300 fatalities every
year [1]. Blind spots are defined as the areas in adjacent
lanes of traffic that are blocked by various structures in the
automobile or other indirect areas not visible to the
driver [2]. Such physical obstructions can potentially cause a
collision if a driver performs a lane change when another
vehicle is present within the blind spot.
To prevent this from happening, some manufacturers have
incorporated blind spot monitoring systems in their vehicles,
i.e. Mercedes Benz (BSM), but these systems are factory
accessories only installed into their vehicles and not
available as an aftermarket product. These blind spot
monitoring systems warn the driver by illuminating a light
and/or producing an acoustic alert. The driver's visual
attention to the road is vital in preventing collisions. Just for
perspective, if a driver is traveling at 70 mph and takes their
eyes off the road for 2 seconds, they would have traveled
205 ft unattended [2]. By using a warning light, it would
require the driver to take their eyes of the road, so the driver
is not aware of the environment in front of them.
In order to help prevent blind spot collisions, a haptic
blind spot alert system has been produced at the University
of South Florida, Tampa, FL, USA. The system utilizes an
aftermarket steering wheel cover with multiple vibrations
motors, an arduino microcontroller, and two ultrasonic
sensors. The alert system detects whether a vehicle is within
the blind spot when the turn signal is engaged. If there is a
vehicle present, a haptic vibration is sent to the driver
through the steering wheel cover.
Related products for blind spot detection systems and
haptic avoidance systems are listed in Section II. The
hardware setup and coding are covered in Section III. The
experimental setup and procedure is covered in detail in
Section IV. The study results are explained in Section V and
finally, the conclusions and future recommendations will be
covered in Section VI and VII.
II. RELATED PRODUCTS AND SYSTEMS
There are multiple products out today that provide blind
spot detection. Volvo offers the "Blind Spot Information
System (BLIS)" on their vehicles as an accessory [3]. BLIS
utilizes a digital camera to monitor the blind spot region and
illuminates a symbol on the mirror housing. Mercedes Benz
offers "Blind Spot Assist" also as an accessory on their S
and CL-Class vehicles [4]. This system uses a short range
radar systems and alerts the driver also by illuminating a
symbol located on the side mirror. Toyota also has a blind
spot detection system (BSM - Blind Spot Monitor) that
works on the same principle as Mercedes and Volvo;
illumination of a symbol on the side mirrors. This requires
the driver not only to check their side mirrors for
approaching vehicles, but also to check for the blind spot
symbol. Figure 1 is an image provided by Toyota of the
blind spot monitor alert on a Toyota Camry.
Figure 1 Toyota Blind Spot Alert
Select automotive companies offer haptic alerts in some
of their vehicles, but none of these systems are for blind spot
detection. Mercedes Benz offers a system called "Lane Keep
Assist", which monitors road markings for any deviation
past a solid line. If the vehicle passes over a solid line, it
provides a vibration to the steering wheel to alert the
driver[4]. This vibration feature is incorporated into the
steering wheel, but is not available for "Blind Spot Assist".
Haptic Blind Spot Alert System
Christopher Nebolsky, Mechanical Engineering Student, University of South Florida
reedkb
Typewritten Text
Spring 2013 Haptics Class Project Paper presented at the University of South Florida, April 29, 2013
At Kettering University in Michigan, David P. Racine has
developed a force feedback steering wheel and gas pedal to
help prevent side collisions caused by blind spots [2].
According to their tests results, side collisions caused by
blind spots were reduced by 47% in a driving simulator
while using his haptic force feedback system. This force
feedback system applies a torque to the steering wheel and a
force to the gas pedal to prevent the driver from making an
improper lane change. This system overrides the control of
the vehicle from the driver, however, many individuals want
to maintain control at all times. In extreme situations, this
system may cause more damage, e.g. if a driver was trying
to avoid a stopped vehicle directly in front of them.
For the haptic alert system presented in this paper,
collisions are prevented by the driver with the assistance of a
haptic response. This haptic response is in the form of a
vibration rather than a force feedback that physically pushes
you from harm's way. This adds an additional sense(touch)
for blind spot recognition, and keeps the drivers attention on
the road
ahead of
them. Another benefit is that this system is universal to any
vehicle since the haptic alert is presented through a steering
wheel cover.
III. HARDWARE SETUP AND CODING
Figure 2 Hardware Setup
The haptic alert system consists of these simple
components:
1. Steering Wheel Cover
2. (4) VPM2 Vibrating Disk Motors
3. Arduino Uno Microcontroller
4. (2) Parallax PING))) Ultrasonic Sensors
A. Steering Wheel Cover Design
The vibration motors are mounted in strategic positions
around the steering wheel cover that accommodate the most
common hand positions. These locations are the standard
"10 and 2", "12 o'clock", and "7 o'clock". The "7 o'clock
position was selected rather than the "5 and 7" due to limited
use of the "5 o'clock" position. Most drivers rest their right
hands on the gear shift or the upper portion of the wheel.
According to the National Highway Traffic Safety
Administration, the conventional "10 and 2" position is not
safe anymore due to new air bag designs [6]. Fortunately,
since the locations of these vibration motors are located all
around the steering wheel, the vibrations can be felt
throughout the circumference of the cover. The VPM2
vibration motors deliver 1G of vibration strength with a 3V
DC supply [7] which is easily sensed and resembles the
vibration of a modern cell phone.
B. Arduino Uno Microcontroller Setup
The Arduino Uno microcontroller was selected to control
the electrical components of the system. All coding for the
Uno was programmed through Arduino's development
environment which uses a simplified version of C/C++
language [8]. The arduino coding uses simple logic to
determine when to execute a haptic alert. First, the
microcontroller checks whether a turn signal has been
enabled. If the turn signal was enabled, it then checks the
ultrasonic sensors for a condition. The condition determines
if there is an object within a certain range from the sensors.
If there is an object present, the microcontroller enables a
haptic alert to the steering wheel. Figure 2 below is the basic
algorithm that the microcontroller has been programmed. A
complete detailed code can be found in the appendix section
of this paper.
C. Ultrasonic Sensors
In order for the alert system to operate, there needs to be a
means of distance observation within the blind spot region.
This is done with the use of ultrasonic sensors which
measures distance with an ultrasonic pulse known as sonar
[9]. The sensor used for this setup was a Parallax PING)))
distance sensor which can measure distances from 3cm to
3m. The ping sensor must be controlled manually by the
microcontroller to by generating a short ping and listening
for an echo to measure distance. Figure 4 is an image
provided by Parallax of how the PING))) operates [9].
Figure 3 Microcontroller Algorithm
Figure 4 PING))) Sensor
The time it takes for the echo to return to the sensor is
directly related to the distance. Generally, a ping of 2
microseconds is pulsed by the sensor. The sound travels at
340 meters per second or 29 microseconds per centimeter.
The total time is then divided by two since the pulse travels
to and from the object within range, and the distance can be
found from the return time. For example, if an object is 2
meters from the sensor, the total time it would take for the
sensor to receive an echo is 11.6 milliseconds. This means
echo took 5.8 milliseconds to travel back from the object. If
we divide 5.8 milliseconds by 29 microseconds, we obtain
200 centimeters or 2 meters. The average highway lane is
3.7 meters wide, and the sensors operating range is 3cm to 3
m. Figure 4 identifies the blind spot range on a standard
vehicle, this image was provided by Mercedes for their BSA
system [4]. The areas are directly perpendicular to the rear
quarter panels of the vehicle. By mounting the sensors on
the rear bumper directly behind the rear wheels, the sensors
can operate correctly and will detect objects that are present
in the adjacent lanes.
Figure 5 Blind Spot Location
For the operation of the haptic alert system, the distance
sensors need to have a form of logic to determine if an
object is within the range of the blind spot. The average
midsize sedan has a width of 1.8 meters, leaving an
estimated 1.9 meters of distance between vehicles on a
highway [10]. If an object was to come within 1.9 meters of
the ultrasonic sensor when a turn signal is engaged, the
microcontroller will alert the driver through the haptic
steering wheel cover.
IV. EXPERIMENTAL SETUP AND PROCEDURE
It was hypothesized that sending a pulsing vibration alert
to an individual will result in a faster alert and response
time. To test this hypothesis, an experimental study program
was generated to evaluate the response time of individuals to
different vibration alerts.
A. Testing Components
The experimental study was conducted with the following
components:
1. Haptic Steering Wheel Cover
2. Arduino Uno Microcontroller
3. Visual Basic Integrated Development Environment
The haptic steering wheel cover was controlled directly by
the arduino microcontroller, and the visual basic program
works in conjunction with the microcontroller by serial port
communication. This allows for a very simple testing
apparatus that does not require many components and
provides a simple and portable testing station.
B. Testing Program
The testing program was generated using visual basic
2010 express edition, which is an event-driven language
development environment. This program presents a user
friendly GUI to the test subjects with simple buttons and a
video screen. The basic operation of the program was to take
basic information from the subject, present a series of haptic
alerts at random times, and record the response times of the
subjects when they pressed a button labeled "Record Time".
Figure 6 Response Time GUI
The data was written to a text file in real time, and was
analyzed by excel for the average time response for each
haptic alert. There were three modes presented to the
subjects at random times throughout the test, each mode was
presented twice for a total of six haptic alerts.
The three modes consisted of:
Table 1Vibration Modes
Mode Loop Duration
Steady Steady for 5 s 5 s
Slow Pulse On for 500 ms, Off for 500 ms 5 s
Fast Pulse On for 200 ms, Off for 50 ms 3 s
The vibrations were controlled by the arduino uno within
its own coding. Once the subject presses the "Begin Test"
button, the visual basic program sends a "1" through the
serial port to the arduino. The arduino was coded to look for
"1" or "0" to begin or end the test. When it receives a "1", it
begins a series of different haptic alerts through the steering
wheel cover.
C. Procedure
Each test subject was informed of what to expect from the
testing program. The GUI presents a video that runs through
the instructions for the study and presents a sample haptic
alert to the subject before the test begins. In order to obtain
accurate response time results, a video of a vehicle traveling
down an interstate was shown to the subject to distract them
visually. Below are the procedures given to each subject:
1. Comfortably seat yourself in front of the computer.
2. Enter age and gender into dialog widow.
3. Grasp the steering wheel with your least dominant
hand.
4. Click "Begin Test"
5. Click "Record Time" every time a haptic alert is
presented
6. When signaled test has completed, click "End
Test".
V. RESULTS
13 subjects participated in the study, ranging from 21 to
56 years old. A total of 78 responses were recorded (13
participants x 3 response time tasks x 2 trials) for haptic
stimuli alerts. Appropriate statistical analysis was performed
by means of ANOVA to study if gender or age had
significant effects on response times.
Analysis of Variance
Source Sum Sq. D.F. Mean Sq. F Prob>F
Subject 1.41994 12 0.11833 9.05 0
Vibration Mode 0.04656 2 0.02328 1.78 0.1902
Error 0.31384 24 0.01308
Total 1.78034 38
After completing the statistical analysis, it was determined
that there is no significant difference between vibrations
modes when it comes to alert response times (F(2,38) = 1.78,
p=0.1902). There were some differences between subjects
and their respected response times (F(12,38) = 9.05, p = 0),
but no factor set them apart from the study group (2 subjects
had slower response times, but varied in age by 30 years).
Figure 7 Subject Comparison
Figure 8 Vibration Mode Comparison
According to a study conducted in [11], the average
response time to visual stimuli is 0.517ms and 0.493ms for
auditory stimuli. The average response time to haptic alerts
in this study was 0.484ms. This shows that the use of a
haptic alerts is just as effective as visual and audio alerts.
However, haptic alerts should not be the only stimuli
presented to a driver when operating a vehicle.
VI. CONCLUSION
With the use of haptic alerts for blind spot monitoring, the
driver has an added sense of security when performing lane
changes. When a driver is operating a motor vehicle on a
highway setting, visual attention is a priority, so to minimize
any unnecessary instances of visual distraction benefits the
driver's safety. The driver can anticipate a haptic alert from
the blind spot alert system before making a visual
verification to ensure a proper lane change with no
collisions. According to the study that was performed, the
response rate to haptic stimuli is just as effective as visual
and audio feedback.
Having this product available to the average consumer
will effectively decrease collisions caused by improper lane
changes. Since the haptic blind spot alert system is universal
to any vehicle, it can be sold as an aftermarket product with
little modification with installation. No longer will
consumers have to purchase specific vehicle brands to obtain
blind spot detection features. The cost of production will be
low due to the simple components used in the system, which
could lead to excellent profitability.
VII. FUTURE RECOMMENDATIONS
This haptic blind spot alert system can be further modified
and enhanced by using different components to amplify the
quality and reliability of the product. Some of these features
could include:
1. Replacing the Ultrasonic Sensor
2. Utilizing System for other Haptic Features
Instead of using ultrasonic sensors to monitor concentrated
areas, a short range radar sensor can be used to monitor the
surrounding area of the vehicle. Ultrasonic sensors are easily
interfered by multiple environmental factors, which could
include rain or dirt. If radar is used in combination with an
effective processor, multiple features can be integrated into
the system. One of these features could include, back up
assist, navigation in poor visual conditions, and collision
prevention. For example, if the driver forgets to signal a lane
change and proceeds to move into a lane occupied by
another vehicle. If the system detects that another vehicle is
within a range of comfort, it can automatically send an alert
to the driver.
Other features could also dictate how the system operates
if certain conditions are met, i.e. vehicle speed or location.
Location can be found under the use of a GPS, and if a
vehicle is on a designated highway or interstate, the system
will know to enable the blind spot alert system. This way the
system does not only depend on vehicle speed since speed
can be very minimal on interstates during rush hours.
REFERENCES
[1] Transportation Statistics Annual Report. Bureau of Transportation Statistics. Web. 9 April 2013. < http://www.rita.dot.gov/bts/sites/rita.dot.gov.bts/files/publications/tran
sportation_statistics_annual_report/index.html> [2] Racine, D. P.; Cramer, N. B.; Zadeh, M.H., "Active blind spot crash
avoidance system: A haptic solution to blind spot collisions," IEEE
International Symposium on Haptic Audio-Visual Environments and Games (HAVE), pp 1-5, 2010.
[3] Hanlen, M., Volvo Launches Blind Spot Information System (BLIS)
Gizmag. Web. 4 April 2013. < http://www.gizmag.com/go/2937/> [4] Mercedes Benz Offers Blind Spot Assist. Mercedes-Benz Dealer